Rapid Determination of Proximates and Amino Acids in Corn Distillers Grains Using Near Infrared Spectroscopy

In swine production the use of corn dry distiller’s grain with solubles (DDGs) in feeds has increased in recent years. With a large number of DDGs suppliers, it is important to understand the nutrient composition of DDGs by supplier especially in relation to amino acids and lysine in particular. Near-infrared (NIR) spectroscopy is an analytical technique used for raw material quality measurement due to its versatility and speed. NIR is a rapid secondary method that relies on an accurate primary reference method for training. In addition, availability of diverse range of samples, as well as application of robust regression algorithms are essential in achieving optimum performance. Hence the objective of this study was to determine the feasibility of measuring proximates and amino acids of DDGs through the use of NIR in un-milled and milled samples. In this study, a wide range of corn DDG samples (N=100), sourced from 15 different plants, were scanned (ground and unground) in two different labs (AB Vista labs in UK and USA) by NIR and results were compared using existing INGOT calibrations (Aunir, UK) for proximate analysis. These samples were also submitted for reference chemistry to two separate labs (Scientec, UK for proximates and University of Missouri for amino acids). New bespoke calibration models were then built for 36 parameters that show significant improvements over existing calibrations. A sub-set of the key results can be seen in the table below in descending order according to RSQ values (RSQ is an indication of agreement between NIR and reference method). The results show that overall the use of NIR can be used to replace traditional reference chemistry to predict proximates and amino acids if a sufficient range of samples are available.

Table 1 – Calibration for a sub-set of the nutrient analyses for DDGs

Nutrient	Mean	SD	SEC	RSQ	SECV	RPD
Starch	4.83	1.97	0.33	0.97	0.38	6.00
Oil (Ether Extract)	6.28	1.33	0.31	0.94	0.36	4.25
NDF (Neutral Detergent Fibre)	30.07	3.22	1.14	0.87	1.39	2.82
Protein	27.60	0.87	0.33	0.85	0.41	2.61
Valine	1.44	0.06	0.03	0.81	0.04	2.30
Moisture	10.80	0.75	0.34	0.80	0.41	2.25
Lysine	1.00	0.06	0.03	0.77	0.04	2.07
Tryptophan	0.21	0.02	0.01	0.75	0.01	1.99
Isoleucine	1.13	0.05	0.03	0.71	0.03	1.85

Abbreviations:

SD: standarad deviation, SEC: standard Error of Calibration, RSQ: R squared, SECV: standard error of cross validation, RPD: ratio of standard error (SECV) to standard Deviation